Abstract- Reconstructing Kinase Network Dynamics to Predict Stochastic Cell Cycle Fate. In a genetically identical and clonally-derived population of cells, stochastic gene expression causes natural cell-to-cell variations in protein expression levels1?8, which causes single cells to exhibit different cell fate when treated with the same stimuli, which can potentially give rise to a population of drug resistant cells2,9, impeding cancer treatment. This phenomenon is referred to as natural phenotypic divergence (NPD), and it arises from how protein expression noise influences the stochastic dynamics of interacting non-linear signaling networks2,9?11. Both protein expression noise and the non-linear nature of signaling dynamics makes it difficult to predict how single cells will respond to a perturbation such as, chemotherapeutics or mitogens. Focusing on cell proliferation responses, we hypothesize that we can predict the timing and probability of cell proliferation by inferring the dynamic connection architecture of the ERK, JNK and Akt signaling networks, which ubiquitously control cell cycle entry. By understanding the connection architecture of these pathways, a causal computational network model can be developed, which can predict the timing and probability of cell proliferation at the single cell level. By combining this network model with live cell imaging experiments spanning different breast cancer subtypes, we can evaluate the generality of how kinase networks control cell cycle entry as well as how cell transformation affects these control systems; which can provide translational insight into novel signaling targets in cancer, predict how transformed cells respond to chemotherapeutics, and the development of transient drug resistance. To achieve this goal, the following aims are proposed: Aim 1. Generate perturbation imaging time course data of ERK, JNK, Akt and S-phase entry dynamics for dynamic network model reconstruction. Aim 2. Construct an ERK-JNK-Akt network model predictive of S-phase entry probability dynamics. Aim 3. Experimentally test model-based predictions of S-phase entry response in a panel of breast cancer cell lines of varying clinical subtypes. Live cell imaging will be used to acquire ERK JNK and Akt dynamics, along with S-phase entry response using kinase translocation reporters (KTR)24,25 and the mCherry-geminin S-phase probe26 respectively. This dynamic data will serve as input data for dynamic modular responses analysis22,23 (DMRA) which will be used to construct a causal network model consisting of the empirical interaction strengths between ERK, JNK and Akt, along with S-phase entry. This network model along with live cell imaging experiments in different breast cancer subtypes will be used to generate and test model predictions, which provide insight the ubiquity of cell cycle entry control systems in mammalian cells and how cell transformation affects that control. The model can provide translational insight into novel cancer signaling pathway targets, as well as predict single cell response to chemotherapeutic and targeted drugs, and consequently transient drug resistance mechanisms.